Contrasting beta diversity among regions: how do classical and multivariate approaches compare?

Abstract

Aim Approaches to calculating beta diversity (β) include classical measures based on alpha (α) and gamma (γ) diversity, and multivariate distance-based measures. Species-area relationships cause measurements of γ to vary, making comparisons of classical β among regions contingent on sampling effort. A recent null-modelling approach has attempted to account for variation in γ by calculating the degree to which β deviates from a random expectation. Here, we clarify the mathematical links between classical and multivariate approaches to measuring β, to derive predictions regarding the reliability of classical, null-model and multivariate approaches. Next, we use four ecological datasets and simulated data to test the consistency of these approaches across sampling effort and γ. We focus on an issue that arises when making comparisons among regions, namely that even small changes to the area sampled can differentially increase measured γ in each region, potentially causing artefacts in β that are driven by methodology rather than biology.Innovation Comparisons among regions using classical and null-model measures change dramatically as sampling effort and γ increase. This change is understood for classical β because of species-area relationships, but not for null-model measures, making comparisons among regions impossible using the null-model approach. Multiple-site dissimilarity shows a similar sensitivity to γ as classical measures. In contrast, pairwise multivariate distances show no systematic effect of sampling effort and γ: increasing the number of sample plots decreases variability but does not alter mean β. Main conclusionsMultivariate pairwise distances are independent of sample size, offering the most robust comparison among regions. The widespread influence of sampling effort and γ indicate that only scale-dependent measures of classical and multiple-site β are comparable, whereas null-model β may not be comparable among regions. However, in cases where γ is well known, multiple-site dissimilarity metrics offer several advantages, and should be strongly considered.

Dispersal enhances beta diversity in nectar microbes

“…These results were consistent with the full dataset and when late June samples were omitted (P < 0.001 in both cases). The complementary metric, pairwise dissimilarity (Bennett & Gilbert 2016), also indicated that flowers exposed to pollinators had a higher average dissimilarity between samples than those from caged and bagged treatments (Tukey HSD P < 0.0001 for all comparisons; Fig S6). Fungal composition varied primarily among months (P < 0.001), but with late June samples omitted, exposed flowers had higher fungal beta diversity than caged flowers (beta diversity P = 0.04, Fig.…”

Dispersal is more important than climate in structuring turtle communities across different biogeographical realms

“…We calculated the beta diversity between pairs of communities (2°9 2°grid cells) for each region following the approach proposed by Baselga (2010). According to current literature, pairwise measures of beta diversity are the best way to avoid biases related to sample size differences when comparing beta diversity between regions (Bennett & Gilbert, 2016). Taxonomic and phylogenetic beta diversity components were calculated using Simpson dissimilarity for the turnover component and a Sorensen-derived formula for the nestedness/richness difference component in the R package 'betapart' (Baselga & Orme, 2012) using the functions beta.pair and phylo.beta.pair, respectively (the equations used in our study are provided in the supplementary material, but see Baselga (2010) for a broad explanation regarding the equations we used for taxonomic beta diversity and Leprieur et al (2012) for explanations of the equations used for phylogenetic beta diversity).…”

Analytical models forβ-diversity and the power-law scaling ofβ-deviation

“…Comparing β -diversity, the variation in species composition, among different regions is an important yet controversial topic in ecology (Kraft et al, 2011; Qian, Chen, Mao, & Ouyang, 2013; Bennett & Gilbert, 2016; Ulrich et al, 2017). A randomization-based null model that was said to correct for the dependence of raw β -diversity on species pool (Kraft et al, 2011) has been widely adopted for this purpose (e.g., De Cáceres et al, 2012; Myers et al, 2013; Vannette & Fukami, 2017; Xing & He, 2019; Zhang, He, Zhang, Zhao, & von Gadow, 2020).…”

“…A randomization-based null model that was said to correct for the dependence of raw β -diversity on species pool (Kraft et al, 2011) has been widely adopted for this purpose (e.g., De Cáceres et al, 2012; Myers et al, 2013; Vannette & Fukami, 2017; Xing & He, 2019; Zhang, He, Zhang, Zhao, & von Gadow, 2020). However, this randomization approach has been criticized on several aspects, including incorrect interpretations of the β -deviation metric and the dependence of the metric on sampling effort and species pool (Qian et al, 2013; Bennett & Gilbert, 2016; Ulrich et al, 2017, 2018). Another limitation of the null model is that it requires individual-level data and thus limits its application to situations when only abundances of individual species in each local community are available.…”

“…Although this link between β -deviation and intraspecific aggregation of species was acknowledged in previous studies (e.g., Crist, Veech, Gering, & Summerville, 2003; Myers et al, 2013), including Kraft et al (2011), this interpretation of β -deviation has not been widely heeded in the literature (e.g., Qian et al 2013). The original interpretation of β -deviation was “a standard effect size of β -diversity deviations from a null model that corrects for γ dependence” (Kraft et al, 2011) but this misinterpretation attracted much criticism on the null model (Qian et al, 2013; Bennett & Gilbert, 2016; Ulrich et al, 2017). To reiterate, β -deviation is neither a measure that corrects for the effect of γ diversity nor a measure to correct for the effect of SAD (Xu, Chen, Liu, & Ma, 2015).…”